Spectrochimica Acta Part A 84 (2011) 275–285

Contents lists available at SciVerse ScienceDirect

Spectrochimica Acta Part A: Molecular and

Biomolecular Spectroscopy
journal homepage: www.elsevier.com/locate/saa

Molecular structure and spectroscopic studies on novel complexes of

coumarin-3-carboxylic acid with Ni(II), Co(II), Zn(II) and Mn(II) ions based on
density functional theory
B.S. Creaven a,b,∗ , M. Devereux c , I. Georgieva d,∗∗ , D. Karcz a,b , M. McCann e , N. Trendafilova d ,
M. Walsh a,b
a
Department of Science, Institute of Technology Tallaght, Dublin 24, Ireland
b
Centre for Applied Science and Health, Institute of Technology Tallaght, Dublin 24, Ireland
c
The Inorganic Pharmaceutical and Biomimetic Research Group, Focas Research Institute, Dublin Institute of Technology, Camden Row, Dublin 8, Ireland
d
Institute of General and Inorganic Chemistry, Bulgarian Academy of Sciences, 11 Acad. G. Bonchev Str., Sofia, Bulgaria
e
Chemistry Department, National University of Ireland, Maynooth Co. Kildare, Ireland

a r t i c l e i n f o a b s t r a c t

Article history: Novel Ni(II), Co(II), Zn(II) and Mn(II) complexes of coumarin-3-carboxylic acid (HCCA) were studied at
Received 14 March 2011 experimental and theoretical levels. The complexes were characterised by elemental analyses, FT-IR,
Received in revised form 1 September 2011 1
H NMR, 13 C NMR and UV–Vis spectroscopy and by magnetic susceptibility measurements. The binding
Accepted 15 September 2011
modes of the ligand and the spin states of the metal complexes were established by means of molecu-
lar modelling of the complexes studied and calculation of their IR, NMR and absorption spectra at
Keywords:
DFT(TDDFT)/B3LYP level. The experimental and calculated data verified high spin Ni(II), Co(II) and Mn(II)
Coumarin
complexes and a bidentate binding through the carboxylic oxygen atoms (CCA2). The model calculations
Metal complexes
DFT modelling
predicted pseudo octahedral trans-[M(CCA2)2 (H2 O)2 ] structures for the Zn(II), Ni(II) and Co(II) complexes
TDDFT and a binuclear [Mn2 (CCA2)4 (H2 O)2 ] structure. Experimental and calculated 1 H, 13 C NMR, IR and UV–Vis
IR data were used to distinguish the two possible bidentate binding modes (CCA1 and CCA2) as well as
NMR mononuclear and binuclear structures of the metal complexes.
© 2011 Elsevier B.V. All rights reserved.

1. Introduction The coordination chemistry and biological activity of coumarin

Coumarin (2H-1-benzopyran-2-one) derivatives are oxygen-
containing heterocyclic compounds that are present in a large
number of natural products [1,2]. The diversity of applications
for coumarin-based compounds has resulted in well established
general methods for their syntheses and still attracts organic and
medicinal chemists towards improvements and modifications of
already existing methods [3–7]. It is also well known, that coumarin
derivatives can yield a wide variety of metal complexes with dif-
ferent coordination modes, spectroscopic properties, and potential
applications. In particular, the complexation of coumarin-derived
ligands to various d-block metal ions has been recognised as a
promising route towards development of new therapeutic agents
[8–13].
∗ Corresponding author at: Department of Science, Institute of Technology Tal-
laght, Dublin 24, Ireland. Tel.: +353 01 404 2889.
∗∗ Corresponding author. Tel.: +359 2 9792592.
E-mail addresses: Bernie.Creaven@ittdublin.ie (B.S. Creaven),
ivelina@svr.igic.bas.bg (I. Georgieva).
derivatives have also been of interest to our group [9,14–16]. In
the past, we have reported a series of Cu(II) and Ag(I) coumarin
carboxylates, which showed promising results in a number of
biological assays [17–19]. For a number of these complexes the
isolation of crystals suitable for X-ray analysis often remained
unsuccessful, and hence their coordination modes were pro-
posed based on experimental data other than X-ray diffraction
[16,20–23]. As a continuation of this work, a number of coumarin-
3-carboxylic acid-derived complexes containing ions other than
Cu(II) and Ag(I) were synthesised. Bidentate coordination through
the lactone carbonylic oxygen and the deprotonated carboxylic
oxygen atoms (CCA1) has been previously proposed for a num-
ber of lanthanide cations (La(III), Ce(III), Nd(III)) with CCA [21,22].
Model calculations of this coordination type, including geome-
try optimization, IR spectra simulations and detailed vibrational
analysis proved the CCA1 binding for the lanthanide ions. An alter-
native coordination of CCA, through the carboxylic oxygen atoms,
CCA2, was predicted for Cu(II) complexes of HCCA [23]. Therefore,
in addition to the available experimental data, DFT modelling of the
complexes was performed. These efforts allowed a confirmation of
the binding mode in the HCCA-derived complexes.

1386-1425/$ – see front matter © 2011 Elsevier B.V. All rights reserved.
doi:10.1016/j.saa.2011.09.041
276 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285
2. Materials and methods
2.1. Physical measurements
All chemicals were purchased from Sigma–Aldrich and used
without purification. Infrared spectra were recorded in the region
of 4000 cm−1 to 400 cm−1 on Nicolet Impact 410 Fourier-Transform
Infrared (FTIR) spectrophotometer using Omnic software. All
infrared spectra were run as KBr discs. Melting point values were
recorded on a Stuart scientific SMP1 melting point apparatus.
Values were taken up to 320 ◦ C. UV–Vis spectra were recorded
on a Hitachi U-2001 spectrophotometer in DMSO (spectopho-
tometric grade) in the region of 260–900 nm. All samples were
dissolved in d6 -DMSO. Signal assignments were made using stan-
dard techniques including DEPT/COSY and CHSHiFt experiments.
Microanalytical data were provided by the Microanalytical Labora-
tory, National University of Ireland, Belfield, Dublin 4, Ireland. Room
temperature magnetic susceptibility measurements were carried
out using a Johnson Matthey magnetic balance in NUI, Maynooth,
Co. Kildare, Ireland. Hg[Co(SCN)4 ] was used as the calibration stan-
dard.
2.2. Computational details
Molecular structure, vibrational (IR), NMR and absorption spec-
tra calculations of M(CCA)2 (H2 O)2 (M = Zn(II), Co(II), Ni(II), Mn(II))
and Mn2 (CCA)4 (H2 O)2 were carried out with DFT/TDDFT method
using Gaussian09 program [24]. The data were compared with
the corresponding results for the isolated neutral ligand (HCCA)
at the same level of theory. Basis set 6-31G(d) (split valence
plus polarization basis set) was applied for the main group ele-
ments: for hydrogen atoms (4s)/[2s], for carbon, and oxygen atoms
(10s4p1d)/[3s2p1d]. To give a better description of the wave func-
tions in the intermolecular region and hence, to satisfy estimation
of the M-CCA and M-OH2 interactions, diffuse functions were
added to the standard basis set (one s and one p set) for the
O atoms as well as for the C1, C2 and C3 atoms (Fig. 1). For
Zn, Co and Ni, the Watchers’ primitive set (14s9p5d) [25] was
employed supplemented with one s, two p and one d diffuse func-
tions [26], and one f polarization function [27]. The final basis set
6-311+G(d) is of the form (15s11p6d1f)/[10s7p4d1f]. Hereafter,
the combined basis set for the metal complex molecule will be
referred as Basis1. Geometry optimization and absorption spectra
calculations of two model complexes of Mn(II), Mn(CCA2)2 (H2 O)2
and Mn2 (CCA2)4 (H2 O)2 , were performed with contracted 6-31G(d)
basis set to prevent time consuming calculations. The computations
using Basis1 and 6-31G(d) basis sets were carried out with the Lee,
Yang and Parr correlation functional (LYP) [28] combined with the
Becke’s non-local three-parameter exchange functional, (B3) [29].
The adequacy of the B3LYP method for prediction of conforma-
tional behaviour, geometrical parameters and vibrational spectrum
of HCCA was proved in our recent investigations [21,22]. Full geom-
etry optimizations were carried out without symmetry constraints.
The minima on the potential energy surfaces were qualified by
the absence of negative eigenvalues in the diagonalized Hessian
matrix, giving imaginary normal vibrational mode. The approxi-
mate assignment of the calculated frequencies to the molecular
vibrational modes was performed using the atom movements in
Cartesian coordinate calculated with the Gaussian program. The
calculated atom movements were further used as an input of the
ChemCraft program which animates them and gives the possi-
bility for a visual inspection of each molecular mode [30]. The
1 H and 13 C NMR chemical shift calculations of HCCA and model
M(CCA)2 (H2 O)2 and Mn2 (CCA)4 (H2 O)2 complexes were performed
in DMSO at B3LYP/6-31+G(d,p) level. The absolute isotropic mag-
netic shielding constants ( i ) were used to obtain the chemical
shifts (ıi =  TMS −  i ) for C and H atoms by referring to the stan-
dard compounds, tetramethylsilane (TMS). The referred compound
TMS was also calculated in DMSO at the same level of theory.
To simulate UV–Vis absorption spectra of Zn(II), Co(II), Ni(II) and
Mn(II) complexes of CCA, 50 lowest singlet excited states were cal-
culated in DMSO by means of TDDFT method and B3LYP/Basis1
level [31]. In spite of difficulties of TDDFT approach in treating
charge-transfer transitions properly [32] it remains a computation-
ally simple and efficient method able to treat practical problems in
a reasonable time scale at low cost as compared to highly corre-
lated ab initio methods [33]. TDDFT calculations have proven to be
quite reliable in describing also the electronic spectra of complexes
containing transition metals in solution [34]. For comparison,
the vertical excitation energies of HCCA and its Zn(II) complex
were calculated at PBE1PBE level [35], recently established as a
more reliable hybrid functional for description of high-energetic
excitations of coumarins [36,37]. NMR and UV–Vis spectra calcu-
lations in DMSO were performed for the optimized geometries
in the same solvent. A non-equilibrium implementation of the
Polarisable Continuum Model (PCM) using the integral equation
formalism variant (IEFPCM) was applied to simulate the solvent
effects [38–40].
3. Results and discussion
3.1. Preparation of complexes
Commercially available coumarin-3-carboxylic acid (HCCA)
(Fig. 1a) was complexed to Zn(II), Co(II), Ni(II) and Mn(II) salts
resulting in the formation of the corresponding complexes,
namely Bis[coumarin-3-carboxylato]zinc(II) dihydrate (1), CCA-
Zn; Bis[coumarin-3-carboxylato]cobalt(II) dihydrate (2), CCA-Co;
Bis[coumarin-3-carboxylato]nickel(II) dihydrate (3), CCA-Ni; and
Tetrakis[coumarin-3-carboxylato]dimanganese(II) dihydrate (4),
CCA-Mn using a modified procedure reported previously by Kar-
aliota et al. [23]. Details are given in the supplementary section.
The addition of 0.02 M NaOH (5 ml) prior to the addition of the
relevant salt, facilitated deprotonation of the carboxylic acid. Com-
plexes were then prepared by addition of aqueous solutions of
the relevant metal salt to DMF solutions of coumarin-3-carboxylic
acid under continuous stirring and refluxing. All the complexes
were easily isolated from the hot reaction mixture by filtration
as the solubility of all the products was significantly lower than
that of free HCCA. Although the formation of binuclear structures
was considered, the elemental analysis data for the zinc, cobalt
and nickel complexes correlated better with a mononuclear struc-
ture, (Table 1). Only the results for the Mn(II) complex 4 were
consistent with the presence of two metal ions. Microanalytical
data supported the presence of water molecules as additional lig-
ands.
3.2. Magnetic susceptibility and spin states of the metal
complexes
As expected, the zinc complex, 1, revealed diamagnetic prop-
erties and did not exhibit a spin moment. The effective magnetic
moment obtained for CCA-Co (2) was 5.47 BM consistent with a
high spin octahedral complex. This value is slightly larger than the
characteristic experimental eff values for high spin Co(II) com-
plexes (4.1–5.2 BM) and the calculated spin-only magnetic moment
(so 3.87 BM) [41]. The higher than expected experimental eff
value might be as a result of a considerable orbital moment s
for 2 which contributes to the overall magnetic moment [42]. eff
values of approximately 5.0 BM are known to be characteristic of
high spin octahedral Co(II) complexes [43]. For CCA-Ni (3) the eff
 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285 277

Fig. 1. DFT/B3LYP/Basis1 optimized structures of: (a) neutral ligand HCCA, (b) trans-M(CCA1)2 (H2 O)2 and (c) trans-M(CCA2)2 (H2 O)2 , [M = Zn(II), Co(II), Ni(II) and Mn(II)].

Table 1
Physical data for metal(II) complexes of HCCA ligand 1–4.

Complex Calc (found) [%] Mw [g/mol] Yield [%] Empirical formula m.p. [◦ C] eff [BM]

C H M*

CCA-Zn (1) 50.07 (50.11) 2.94 (2.92) 13.63 (13.96) 479.71 73 C20 H10 O8 Zn·2H2 O 206–208 –
CCA-Co (2) 50.76 (50.66) 2.98 (3.02) 12.45 (–) 473.25 87 C20 H10 O8 Co·2H2 O 254–256 5.47
CCA-Ni (3) 50.78 (50.64) 2.98 (3.01) 12.41 (12.78) 473.01 85 C20 H10 O8 Ni·2H2 O >300 3.42
CCA-Mn (4) 54.07 (54.16) 2.95 (2.89) 12.37 (–) 888.50 81 C40 H20 O16 Mn2 ·2H2 O >300 7.38

* = Metal, – = not determined.

278 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285
value was 3.42 BM which correlates well with the typical eff val-
ues for high spin Ni(II) octahedral complexes (2.9–3.4 BM) [41]
(calc. so 2.83 BM) and it is similar to values reported for other
mononuclear, deformed octahedral Ni(II) complexes containing
carboxylate-substituted ligands (3.04–3.28 BM) [44,45]. CCA-Mn
(4) exhibited an effective magnetic moment value of 7.38 BM and
was consistent with high-spin octahedral Mn(II) complexes. The
value is much higher than the values characteristic for high-spin
mononuclear structures (∼5.9 BM, calc. so 5.92 BM) [41] and is
indicative of a high-spin binuclear complex. A similar value was
reported for a binuclear Mn(II) structure with bridging carboxylate
ligands reported by Kiskin et al., eff = 7.0 BM at a temperature of
300 K [46].
3.3. Molecular modelling of Zn(II), Co(II), Ni(II) and Mn(II)
complexes of CCA in the gas phase
Molecular electrostatic potential calculations of the active
deprotonated form of the ligand (CCA) suggested two possible
sites of bidentate binding to the metal ion: first through the car-
bonyl and deprotonated carboxylic oxygens (CCA1), and second
through both the carboxylic oxygens (CCA2) [21]. All possible spin
states of Co(II), Ni(II) and Mn(II) complexes were calculated for the
model [MCCA1]+ and [MCCA2]+ species and their relative electronic
energies are given in Table S1 (Supplementary material). The cal-
culated energies of the models studied predicted a stabilization of
the [MCCA2]+ species (bidentate binding through the carboxylic
oxygens) in the high spin states, namely a triplet state for Ni(II), a
quartet state for Co(II), and a sextet state for the Mn(II) complex.
The results obtained are in very good agreement with the metal
spin state predictions based on the effective magnetic moments
found (eff ). Furthermore, all calculations were performed for the
high spin state complexes.
The CCA1 and CCA2 coordination modes were also exami-
ned for the [M(CCA)2 (H2 O)2 ] complexes as suggested by the
elemental analysis, M = Zn(II), Co(II) and Ni(II) (Fig. 1b and c).
Both cis- and trans-CCA isomers were considered. Geometry
optimizations of the model complexes showed several energy
minima structures with difference of 2–3 kcal/mol. The relative
energies are given in Table S2 (Supplementary material). The opti-
mized [M(CCA1)2 (H2 O)2 ] complexes (bidentate CCA coordination
through carboxylic and lactone-carbonyl oxygens) show pseudo
octahedral structures and the trans-CCA1 position is the preferred
one (Fig. 1b). The bidentate CCA2 binding via both carboxylic oxy-
gens leads to more stable structures compared to that of CCA1.
The trans-[M(CCA2)2 (H2 O)2 ] complexes reveal deformed octahe-
dral structures stabilized by two Hw · · ·O hydrogen bonds formed
between the coordinated water hydrogens and coumarin carbonyl
atoms (Fig. 1c). In contrast to the trans-[M(CCA1)2 (H2 O)2 ] struc-
ture the aromatic ring of the trans-[M(CCA2)2 (H2 O)2 ] structure
is perpendicular to the C3O3O4M plane. The model calculations
predicted another energy minimum structure stabilized by four
intramolecular O· · ·Hw hydrogen bonds and tending to tetrahe-
dral metal coordination with monodentate CCA binding via the
carboxylic oxygen; however, such a tetrahedral coordination is in
conflict with the effective magnetic moment values for Ni(II) and
Co(II) and their predicted octahedral coordination modes. More-
over, the calculated IR spectrum of the tetrahedral monodentate
[M(CCA)2 (H2 O)2 ] complex shows an intense band at 1313 cm−1 due
to stretching ␯(C3O3) vibration, which is not observed in the exper-
imental IR spectrum of the CCA-M(II) complexes. Therefore, for
Zn(II), Ni(II) and Co(II) complexes we accepted bidentate binding of
CCA via the carboxylic oxygens in the trans position and octahedral
complex structures, [M(CCA2)2 (H2 O)2 ] (Fig. 1c). In order to con-
firm this hypothesis, the IR, NMR and UV–Vis absorption spectra at
the optimized octahedral structures were calculated and compared
to those obtained experimentally. Comparisons of structures with
CCA1 binding (trans-[M(CCA1)2 (H2 O)2 ]) and their spectroscopic
properties were also carried out. These efforts allowed determi-
nation of the spectroscopic properties specific to CCA1 and CCA2
binding modes in the metal complexes 1–3.
The elemental analysis and effective magnetic moment value
of 7.38 BM suggested a binuclear structure for the Mn(II) com-
plex of CCA. At the same time, the IR spectrum of the Mn(II)
complex is very similar to those of the mononuclear Zn(II), Co(II)
and Ni(II) complexes of CCA. Therefore, both mononuclear trans-
[Mn(CCA2)2 (H2 O)2 ] and binuclear [Mn2 (CCA2)4 (H2 O)2 ] complexes
were considered and their IR, NMR and absorption spectra were cal-
culated and compared with the available spectroscopic data. Both
model structures possess minima in their potential energy surface
and both are stabilized via H-bonds formed between the coor-
dinated water hydrogens and coumarin lactone-carbonyl atoms
(Figs. 1c and 2).
3.4. Experimental and theoretical NMR spectra
The experimental and calculated 1 H and 13 C NMR chemical
shifts of HCCA and its [Zn(CCA)2 (H2 O)2 ] complex in DMSO are
presented in Table 2. The trend in the experimental chemical
shift in HCCA and its Zn(II) complex was followed and compared
with the calculated ones for model HCCA, trans-[Zn(CCA1)2 (H2 O)2 ]
and trans-[Zn(CCA2)2 (H2 O)2 ]. A comparative NMR analysis gives
additional evidence for the complex formation and the ligand coor-
dination mode.
The measured 1 H NMR spectrum of HCCA and its correspon-
ding Zn(II) complex revealed five signals in the aromatic region.
Four of these signals were assigned to the benzene ring of the
coumarin moiety. The fifth peak in this region was attributed to the
vinyl H4 proton. Upon complexation the positioning of this peak
shifted upfield, from 8.73 to 8.47 ppm. The 1 H NMR calculations
revealed that the positioning of the H4 signal varied for trans-
[Zn(CCA1)2 (H2 O)2 ] and trans-[Zn(CCA2)2 (H2 O)2 ] complexes. The
value of 9.04 ppm calculated for the free HCCA ligand, changed to
7.97 ppm on simulation of the trans-[Zn(CCA2)2 (H2 O)2 ] formation,
while for the trans-[Zn(CCA1)2 (H2 O)2 ] the positioning of H4 was
calculated as 9.47 ppm. The calculated upfield shift of H4 for trans-
[Zn(CCA2)2 (H2 O)2 ] complex was in agreement with the acquired
data. This might be evidence for the actual coordination mode of
the complex.
The calculated chemical shift value for the carboxylic proton
was 12.96 ppm. This signal was not observed in the acquired HCCA
spectrum, probably as a result of rapid exchange of the carboxylic
proton with protons of residual water in the DMSO solvent. All
the remaining signals either did not change their positioning upon
complexation or the changes in their shifts were minimal.
The 13 C NMR data for HCCA and its Zn(II) complex were
also compared to their corresponding calculated values. In the
HCCA spectrum, the signal at 148.3 ppm was assigned to the
vinyl C4 carbon. The positioning of this peak seemed very sen-
sitive to environmental changes and might be particularly useful
in the determination of the coordination mode of the complex.
The acquired data showed a migration of the C4 signal upon
complexation to 144.7 ppm. According to the calculations this
result indicated the CCA2 coordination of the complex (ligand
calc. 151.5 ppm → complex calc. 137.5 ppm). Calculated values for
CCA1 binding mode resulted in a downfield shift of the C4 sig-
nal (ligand calc. 151.5 ppm → complex calc. 152.9 ppm). Another
significant downfield shift upon complexation was observed for
the C2 carbon (116.2 ppm → 124.4 ppm). The corresponding calcu-
lations supported the CCA2 coordination mode. In the 13 C NMR
spectrum of HCCA, the C1 signal was visible at 164.1 ppm and the
C3 signal – at 156 ppm. Upon complexation, the C1 signal was
 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285 279

Fig. 2. B3LYP/6-31G(d) optimized structure of Mn2 (CCA2)4 (H2 O)2 .

shifted upfield (157.8 ppm), whereas the C3 signal was shifted
downfield (167.4 ppm). Changes in the positioning of these sig-
nals correlated with their corresponding calculated results for the
trans-[Zn(CCA2)2 (H2 O)2 ]. Overall, the combined acquisition and
calculated NMR results for HCCA and its Zn(II) complex gave a
strong indication of the CCA2 coordination mode of the complex.
The calculated NMR results for trans-[Ni(CCA1/CCA2)2 (H2 O)2 ]
and trans-[Co(CCA1/CCA2)2 (H2 O)2 ] complexes showed similar
trends in the chemical shifts upon complexation as those of
the Zn(II) complex and they are given in Table S3 (Supplemen-
tary material). Comparison of the calculated 13 C NMR data for
[Mn(CCA2)2 (H2 O)2 ] and [Mn2 (CCA2)4 (H2 O)2 ] complexes indicated
a larger downfield positioning of C2 and C3 atoms and a larger
upfield shift of C4 atom for the first complex, Table S4 (Supple-
mentary material). The H4 signal did not change in the binuclear
Mn(II) complex, whereas it moved to higher field in mononuclear
[Mn(CCA2)2 (H2 O)2 ] complex. Based on these results it was possible
to distinguish between mono- and binuclear Mn(II) complex.

Table 2
Calculated 1 H and 13 C NMR chemical shifts (ppm) in DMSO at B3LYP/6-31+G(d,p) level for HCCA ligand and its Zn(II) complex compared to the experimental data in DMSO.

Atom HCCA Zn(CCA)2 (H2 O)2

Exp Calc Exp CCA2 calc CCA1 calc

a
H (C4) 8.73 9.04 8.47 7.97 9.47
H (C6) 7.88d 8.01 7.84d 7.81 8.04
H (C9) 7.88d 7.79 7.84d 7.88 7.82
H (C7) 7.71t 8.22 7.64t 7.67 7.75
H (C8) 7.71t 7.73 7.64t 7.92 8.06
H(O3) – 12.96 – – –
H(O)w – 2.80 2.57, 2.46
H(O)w – 6.54 (HB)
C1 164.1 162.6 157.8 160.4 163.2
C3 156.7 161.9 167.4 175.8 163.4
C10 154.6 153.5 154.0 153.7 154.3
C4 148.3 151.5 144.7 137.5 152.9
C8 134.4 134.8 132.7 130.8 133.7
C6 130.3 130.1 129.3 129.7 129.3
C7 124.9 123.7 123.4 123.4 123.3
C5 118.4 119.5 118.5 117.6 119.1
C9 117.4 114.6 115.8 114.9 113.6
C2 116.2 112.4 124.4 124.7 116.4
a
Atom numbering is given in Fig. 1.
280 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285

Fig. 3. TDDFT/B3LYP/Basis1 absorption spectra of free HCCA and trans-M(CCA2)2 (H2 O)2 complexes in DMSO, M = Ni(II), Co(II) and Zn(II).
 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285
Fig. 4. Calculated absorption spectra of Mn2 (CCA2)4 (H2 O)2 and Mn(CCA2)2 (H2 O)2 complexes at B3LYP/6-31G(d) level.
The 1 H and 13 C NMR spectra for Ni(II), Co(II) and Mn(II) com-
plexes were not acquired due to the strong paramagnetic character
of these compounds.
3.5. Experimental and theoretical UV–Vis spectra
The measured UV–Vis spectra of HCCA and its Ni(II), Co(II) and
Zn(II) complexes were analyzed on the basis of TDDFT calcula-
tions of the vertical excitation energies in DMSO (Supplementary
material, Table S5). The calculated absorption spectra and the
character of the transitions for HCCA and trans-[M(CCA2)2 (H2 O)2 ]
complexes (M = Ni(II), Co(II) and Zn(II)) are given in Fig. 3. The
experimental absorption spectrum of HCCA revealed two bands
with maxima at 329 nm (low intensity) and 285 nm respectively.
The calculated absorption maxima at 329 nm and 308 nm corre-
late with the experimental data and correspond to the vertical
excitation energies of the first 1 A (␲␲*) and second 1 A (␲␲*)
excited states, respectively (Fig. 3). According to the calculations
␲ → ␲* transitions with large oscillator strengths are expected
also at 221 nm and 209 nm in the UV–Vis spectrum of HCCA
(Supplementary material, Table S5).
The calculated B3LYP energy of the second 1 A (␲␲*) excited
state of HCCA is lower by 0.32 eV than the experimental one
281
in DMSO solution. Such underestimation of the transition ener-
gies at B3LYP level is known from previous studies of coumarins
[36,37,47]. TDDFT/PBE1PBE/Basis1 calculations of HCCA and trans-
[Zn(CCA2)2 (H2 O)2 ] in DMSO showed higher excitation energies by
0.1–0.15 eV as compared to the B3LYP data, in line with recent
theoretical studies [36,37]. In case of HCCA, PBE1PBE vertical excita-
tion energies better approach the experiment (than those at B3LYP
level), however the second 1 A (␲␲*) excited state remains in being
lower by ∼0.2 eV. In the case of the Zn(II) complex however, the
PBE1PBE results more deviate from the experimental excitation
energies (as compared to the B3LYP data). Hence, the B3LYP/Basis1
excitation energies of the metal complexes of HCCA are discussed
below.
For the trans-[Ni(CCA2)2 (H2 O)2 ] complex, the calculated verti-
cal excitation energies at 319 nm and 292 nm with large oscillator
strengths are consistent with the observed absorption maxima at
311 nm and 297 nm. This result confirmed the structure suspected.
According to Mulliken population analysis, the excitation related
to the first absorption band at 311 (exp.)/319 (calc.) nm involves
transition from CCA(␲) orbital to a mixed orbital with Ni(dxz ) and
CCA(␲*) contributions. Therefore, this band has ligand-to-mixed
metal/ligand charge transfer character and it appears characteris-
tic for the [Ni(CCA2)2 (H2 O)2 ] complex (Fig. 3). The second band at
282 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285
Fig. 5. Calculated (scaled by factor 0.951) IR spectra of HCCA, Zn(CCA1)2 (H2 O)2 and Zn(CCA2)2 (H2 O)2 complexes at B3LYP/Basis1 level (a) and of Mn2 (CCA2)4 (H2 O)2 and
Mn(CCA2)2 (H2 O)2 at B3LYP/6-31G(d) level.
297 (exp.)/292 (calc.) nm reveals CCA(␲) → CCA(␲*) character. The
d-d transitions are calculated in the range 890–500 nm with very
low intensity.
For the complex trans-[Co(CCA2)2 (H2 O)2 ] the calculated
absorption bands at 317, 305, 295, 292 nm respectively, correlate
well with the observed UV–Vis bands at 320, 308, 297, 294 nm
(Fig. 3 and Table S5). According to the population analysis of
the model Co(II) complex, bands at 317, 305 and 295 nm are
attributed to metal–ligand charge transfer transitions. The first
band represents a mixed Co(dxz )/CCA(␲) orbital to a CCA(␲*) orbital
transition, while the second and the third ones represent a CCA(␲)
to a Co(dyz )/CCA(␲*) orbital transitions. Bands at 305 nm and
295 nm reveal significant intraligand charge transfer contribution,
CCA(␲) → CCA(␲*).
TDDFT calculations of the vertical excitation energies for
trans-[Zn(CCA2)2 (H2 O)2 ] complex predict bands at 309, 298, and
292 nm with CCA(␲) → CCA(␲*) transition character (Fig. 3). These
bands are related to the observed absorption bands at 322,
297, and 291 nm respectively. d-Contribution for these transi-
tions was not calculated. The intense band at 202 nm reveals
CCA(␲) → Zn(s)/CCA(␲*) character.
The absorption spectra of the two model Mn(II) complexes
were simulated by TDDFT calculations (Fig. 4 and Table S6). For
the mononuclear complex of Mn(II), [Mn(CCA2)2 (H2 O)2 ], a band
at 302 nm with CCA(␲) → CCA(␲*) character was calculated. For
[Mn2 (CCA2)4 (H2 O)2 ] a complex high intense absorption band at
321 nm was calculated and assigned to a ligand-to-ligand charge
transfer transition. The 321 nm value better corresponds to the
band observed at 327 nm and supports the formation of a binuclear
CCA complex of Mn(II). The second high intensity band calculated at
295 nm for the binuclear and at 285 nm for the mononuclear Mn(II)
complex of CCA revealed metal-to-ligand charge transfer character,
it is due to a transition from a mixed Mn(dxz )/CCA(␲) orbital to a
CCA(␲*) orbital. Unfortunately, bands at these wavelengths were
not observed in the experimental UV–Vis spectrum.
3.6. Experimental and theoretical IR spectra
Geometry optimization and IR spectrum calculations were per-
formed for the trans-[Zn(CCA2)2 (H2 O)2 ] complex. The calculated
and scaled by factor 0.951 IR spectra in the range 200–1800 cm−1 of
HCCA, trans-[Zn(CCA1)2 (H2 O)2 ] and trans-[Zn(CCA2)2 (H2 O)2 ] are
shown in Fig. 5a. Selected experimental and calculated IR data and
assignments of HCCA and both trans-[Zn(CCA)2 (H2 O)2 ] model com-
plexes are given in Table 3.
The strong IR band observed at 3200 cm−1 in the IR spectrum
of CCA-Zn was assigned to a ␯(OH) vibration. Different ␯(OH) band
positioning was calculated for CCA1 and CCA2 coordination modes
and this finding can be used to distinguish the binding mode in
the metal complexes of CCA. The calculated geometry of trans-
[Zn(CCA2)2 (H2 O)2 ] revealed H-bond formation Hw · · ·O2 (Fig. 1c)
resulting in elongation of O–H bond lengths. Therefore, ␯(OH) fre-
quency was expected to shift to lower frequencies compared to
its positioning in [Zn(CCA1)2 (H2 O)2 ] complex where intramolec-
ular H-bonding was not present (Fig. 1b). Thus, the calculated
␯as (OH) for trans-[Zn(CCA2)2 (H2 O)2 ] was obtained at lower fre-
quencies (3316 cm−1 ) in better agreement with the experimental
data than in trans-[Zn(CCA1)2 (H2 O)2 ] (3452 cm−1 ). The experimen-
tal IR spectra of the Ln(III)-CCA complexes [22], where the CCA1
binding is operative, showed a ␯(OH) band at ∼ 3430 cm−1 , whereas
the experimental IR spectra of the M(II)-CCA complexes (M = Zn, Co,
 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285 283

Table 3
Selected experimental and calculateda IR frequencies (in cm−1 ) and IR intensities (I, in km mol−1 ) of HCCA, Zn(CCA2)2 (H2 O)2 and Zn(CCA1)2 (H2 O)2 complexes.

HCCA Zn compl Zn(CCA2)2 (H2 O)2 Assignment Zn(CCA1)2 (H2 O)2

Exp. Calc. Exp. Calc. Calc.

Freq. I Freq. I Freq. I

Overlap 3607 275 ␯as (OH)w 3614 166

3607 0 ␯s (OH)w 3614 7
– – – 3200s 3323 1931 ␯as (OH)w 3469 176
3323 25 ␯s (OH)w 3468 5
3056s 3243 407 – – – ␯(O–H)coum – –
1745vs 1729 447 – – – ␯(C3 O4)fr carbox 1649 779
1682vs 1657 552 1667s 1663 1760 ␯(C1 O2)carbonyl 1576 972
1616s 1633 193 ␦(OHO) 1620 421
1609s 1583 264 1600m,d 1603 81 ␯(CC) + ␦(CCH) 1607 253
1568s 1533 180 1585s 1583 236 ␯(CC) + ␦(CCH) 1569 52
– – – 1561s 1549 596 ␯as (CO2 ) + ␦(CCO) – –
1489 1459 10 1457s 1434 111 ␯(CC) + ␦(CCH) 1529 388
1458 14
1422 43
1452s 1374 254 – – ␦(COH) + ␯(O· · ·H) – –
1421sh
– – – 1396s 1389 482 ␯s (CO2 ) + ␯(CC) – –
1373s 1335 23 – – ␯(CC) + ␦(CCH) 1328 15
– – 1288m – – ␯(C3O3) 1282 720
1298w 1318 1 1288m 1331 32 ␦(CCH) + ␯(CC) – –
1244w 1260 21 1259m 1250 108 ␯(CC) + ␦(CCH), 1247 361
1226 0 1221m 1214 66 ␯(CC) + ␦(CCH) 1217 153
1227m 1194 56 – – – ␯(C10O1) 1188 8
1206m 1176 137 1184m 1146 97 ␦(CCH) + ␯(C–Olac ) 1138 130
1164m 1136 16 1153m 1121 31 ␦(CCH) + ␯(C–Olac ) 1115 130
1121m 1106 16 1132m 1095 21 ␦(CCH) 1095 18
1041s 986 54 1046m 1014 168 ␯(C1O1) 995 44
989s 959 13 1024w 1005 19 ␦(CCH)o.o.p. 1005 13
963w 950 3 974w 945 0 ␦(CCH)o.o.p. 991 55
938m 914 1 929w 921 32 ␦(CCH)o.o.p. 955 20
924m 887 5 879m 889 15 ␦(CCC)ip – –
882m 841 3 – – – ␦(CCH)o.o.p. – –
832m 791 74 – – – ␦(COH)o.o.p. – –
810m 795 312 ␦(OHO) ␦(ZnOHHB ) – –
801m 751 50 779s 767 129 t(OCCO2 )ip /Zn-O3 753 128
741m 749 166 t(CCCO2 )
770m 739 55 711w 736 121 ␦(CCH)o.o.p.
744w 723 5 646s 696 14 ␦(CCC)ip
623w 590 65 ␦(HOH) 593 424
571 214
572 79 Zn-Oc
593m 546 243 ␦ (ZnOH)
587w 619 12 ␯(O· · ·H)
574w 566 9 577m 538 28 Ringip
473m 521 30 ␦(CCO)ip /␦ (HOH) 472 519
470s 450 15 463m 504 8 ringip
381 243 ␯as (Zn-O), t(HOPtOH) 457 38
261 120 Zn-Ow 218 28
a
Calculations at DFT/B3LYP/6-31G(d) level, frequencies are scaled by factor 0.951.

Ni, Mn) with suggested CCA2 binding (this work) exhibited a ␯(OH)
band at about 3200 cm−1 .
According to the calculations, the strong band observed at
1745 cm−1 in the IR spectrum of the free HCCA ligand is assigned
to the carboxylic ␯(C3 O4) mode (calc. 1729 cm−1 ). The calcu-
lations showed that the coordination of Zn(II) via the carboxylic
CO2 − group produces equalization of C3 O4 and C3–O3 bonds
(elongation of the double C3 O4 bond, ∼0.055 Å) and signifi-
cant lowering of the asymmetric ␯as (CO2 ) frequencies ( ∼180,
209 cm−1 ). It should be noted that in HCCA, distinct ␯(C3 O4)
and ␯(C3–O3) carboxylic vibrations are observed, whereas in trans-
Zn(CCA2)2 (H2 O)2 they appear as ␯as (CO2 ) and ␯s (CO2 ) vibrations.
The calculated ␯as (CO2 ) frequencies at 1549, 1520 cm−1 are related
to the strong experimental band at 1560 cm−1 . Thus, the theoreti-
cally predicted shift of ca. 195 cm−1 approaches the experimental
one, 184 cm−1 : ␯(C3 O4) 1745 cm−1 → ␯as (CO2 ) 1561 cm−1 . The
␯s (CO2 ) frequency in the Zn(II) complex is calculated at 1390
and 1389 cm−1 and thereby the strong band at 1396 cm−1 in
the Zn(II) complex spectrum is assigned to ␯s (CO2 ). The experi-
mental peak at 1396 cm−1 shows splitting due to the presence
of two CCA ligands and this finding confirms the assignment.
It is worth mentioning that this band is missing in the calcu-
lated IR spectrum of [Zn(CCA1)2 (H2 O)2 ] complex and therefore it
could be considered as characteristic of CCA2 coordination. The
theoretically predicted ␯(OCO) = ␯as (CO2 ) − ␯s (CO2 ) difference is
160 cm−1 . Accepting the assignment of the experimental bands
mentioned above, the experimental ␯as (CO2 ) − ␯s (CO2 ) difference
is estimated of 165 cm−1 (1561 − 1396 cm−1 ) and it is in the range
of the predicted values. Unfortunately, the magnitudes obtained
are characteristic not only for bidentate binding of the CO2 − group,
they are reported in the literature also for bridging coordination of
the CO2 − group [48].
The vibrational behaviour of the lactone carbonylic moiety,
␯(C1 O2) was also considered. In the free HCCA ligand, the car-
bonylic group is included in intramolecular hydrogen bonding
(Fig. 1a) and the ␯(C1 O2) frequency is slightly lowered: at
284 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285
1657 cm−1 (calc.)/1682 cm−1 (exp.). In the [Zn(CCA2)2 (H2 O)2 ] com-
plex, the lactone carbonylic group does not bind to the metal ion
but it is included in hydrogen bonding, in this case with the coor-
dinated H2 O molecules, Fig. 1c. Thus, the ␯(C1 O2) frequencies in
the complex should be close to the corresponding one of the ligand,
1657 cm−1 . This was the case of the Zn(II) complex, the calculated
␯(C1 O2) frequency is 1663 cm−1 (Table 3) and it indicates that the
strong band observed at 1667 cm−1 in the complex spectrum is due
to the lactone carbonylic stretching mode. The ligand → complex
band shift is small, only 19 cm−1 and could not indicate coordina-
tion via the lactone carbonylic group (along with the deprotonated
carboxylic oxygen).
At the same time, for the alternative coordination mode,
CCA1 via the carboxylic and carbonylic oxygens (Fig. 1b), the
␯(C3 O4) frequency of the uncoordinated group was predicted at
1649 cm−1 . As compared to the ligand, the calculated ␯(C3 O4)
shift is 80 cm−1 (av.): 1729 cm−1 (ligand) → 1649 cm−1 (IR active,
for [Zn(CCA1)2 (H2 O)2 ] complex). This shift value is much smaller
in comparison with that calculated for the coordinated C3–O4
group in CCA2 coordination (av. 195 cm−1 ). In agreement with
the calculated IR spectra of the two Zn(II) complexes, the
band observed at 1667 cm−1 could be either due to the car-
bonylic stretching mode, ␯(C1 O2) of CCA2 coordination, or due
to the carboxylic stretching mode, ␯(C3 O4) of CCA1 coordi-
nation type. In the free CCA ligand, the double C1 O2 and
C3 O4 bonds differ and the ␯(C1 O2) vibration appears at
higher frequencies with ∼90 cm−1 . In trans-[Zn(CCA1)2 (H2 O)2 ] and
trans-[Zn(CCA2)2 (H2 O)2 ] (Fig. 1b,c), however the corresponding
uncoordinated C3 O4 and C1 O2 frequencies are very similar due
to the H-bonding of C1 O2 group and its frequency lowering. Thus,
one may conclude that the IR spectral behaviour of the stretching
carbonylic frequency of the ligand, ␯(C1 O2) and the metal com-
plex ␯(C1 O2)/␯(C3 O4) is not characteristic and could not be used
to elucidate the CCA coordination mode.
According to the calculations and in agreement with other
assignments in the literature, the band at 1616 cm−1 in the com-
plex spectrum is assigned to the characteristic bending modes of
the coordinated water molecules [48].
The experimental IR spectra of Ni(II), Co(II) and Zn(II) com-
plexes are very similar and therefore we believe that the vibrational
assignments made for Zn(II) complex will be valid for Ni(II) and
Co(II) complexes of CCA as well.
The calculated IR spectra of [Mn(CCA2)2 (H2 O)2 ] and
[Mn2 (CCA2)4 (H2 O)2 ] are quite similar, Fig. 5b. Selective vibrational
frequencies are given in Table S7 (Supplementary material). At
first sight, the carbonyl ␯(C1 O2) vibrations in the simulated IR
spectra above differ slightly (1693, 1732 cm−1 ). It should envisage
that in the complexes studied the carbonyl C1 O2 group is not
bound to Mn(II) but it is involved in hydrogen bonding. Hence,
the carbonyl ␯(C1 O2) vibrations appear sensitive to the H-bond
strengths, which are evaluated by DFT calculations. In conclusion,
the calculations predict that the IR spectroscopic data could not
significantly help to distinguish mono- and binuclear Mn(II) com-
plexes with the same ligand binding mode. The result obtained
explains the similar IR spectra of the metal complexes studied
although the large effective magnetic moment for Mn(II) complex
of 7.38 BM suggests a binuclear complex.
4. Conclusions
Novel Zn(II), Co(II), Ni(II) and Mn(II) complexes of coumarin-
3-carboxylic acid were synthesized and characterised by FTIR,
1 H NMR, 13 C NMR and UV–Vis spectroscopies and by DFT/B3LYP
calculations. Mononuclear diamagnetic Zn(II), quartet spin state
Co(II) and triplet spin state Ni(II) complexes, [M(CCA)2 (H2 O)2 ] and
binuclear sextet spin state Mn(II) complex, [Mn2 (CCA)4 (H2 O)2 ],
were suggested by elemental analysis and magnetic susceptibility
measurements. Energy calculations of model CCA-M(II) complexes
with the possible spin states confirmed the stabilization of the
high spin Ni(II), Co(II) and Mn(II) complexes. Two bidentate lig-
and binding modes, CCA1 (through carbonylic and deprotonated
carboxylic oxygens) and CCA2 (through both carboxylic oxygens)
were examined in the model complexes by means of a compari-
son of their calculated and experimental IR, NMR and absorption
spectra. The theoretical and experimental spectroscopic data are in
better agreement with the mononuclear trans-[M(CCA2)2 (H2 O)2 ]
models for Co(II), Zn(II), Ni(II) and binuclear [Mn2 (CCA2)4 (H2 O)2 ]
model. The bidentate CCA binding through the carboxylic oxygens
in trans-position and the suggested pseudo octahedral structures
were theoretically confirmed. The comparative 13 C and 1 H NMR
data analysis of HCCA and its Zn(II), Co(II), Ni(II) and Mn(II)
complexes showed that the 1 H and 13 C signal behaviour could
distinguish the CCA1 and CCA2 bidentate binding manners as
well as the mononuclear and binuclear Mn(II)complexes. ␯(OH)
frequencies of mononuclear trans-[M(CCA)2 (H2 O)2 ] and binuclear
[Mn2 (CCA)4 (H2 O)2 ] complexes with CCA2-M binding are observed
and calculated at lower frequencies ∼3200 cm−1 due to intra- or
intermolecular O· · ·Hw bonds. In CCA1-M(II) complexes ␯(OH)w
appear at ∼3400 cm−1 . The strong bands observed at 1561 and
1389 cm−1 in the IR spectra of the CCA-M(II) complexes assigned
to ␯as (CO2 ) and ␯s (CO2 ) modes explicitly reveal the CCA2 manner
of binding in the M(II) complexes.
Acknowledgements
This research was supported by the Technological Sector
Research Programme, Strand 1, under the European Social Fund.
D.K. gratefully acknowledges the EMBARK fellowship for addi-
tional funding. The 13 C NMR analysis for compound 1 was carried
out with the kind assistance of Dr Brian Murray, using Avance
III 400 MHz, courtesy of Bruker Biospin, UK. I.G. and N.T. thank
the National Science Fund of Bulgaria for the financial support
(Grant Nos. DO-02-233/2008 and DCVP-02/2/2009). The calcula-
tions were performed on the computer cluster MADARA (Project
RNF01/0110/2008 NSF of Bulgaria).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.saa.2011.09.041.
References
[1] H.E. Kleiner, X. Xia, J. Sonoda, J. Zhang, E. Pontius, J. Abey, R.M. Evans, D.D. Moore,
J. DiGiovanni, Toxicol. Appl. Pharm. 232 (2008) 337.
[2] T. Smyth, V.N. Ramachandran, W.F. Smyth, Int. J. Antimicrob. Agents 33 (2009)
421.
[3] F. Borges, F. Roleira, M. Milhazes, L. Santana, E. Uriate, Curr. Med. Chem. 12
(2005) 887.
[4] D.S. Bose, A.P. Rudradas, M.H. Babu, Tetrahedron Lett. 43 (2002) 9195.
[5] B. Tyagi, M.K. Mishra, R.V. Jasra, J. Mol. Catal. 276 (2007) 47.
[6] J. Jayashankaran, R.D.R.S. Manian, R. Raghunathan, Tetrahedron Lett. 47 (2006)
2265.
[7] D. Maes, S. Vervisch, S. Debenedetti, C. Davio, S. Mangelinck, N. Giubellina, N.
De Kimpe, Tetrahedron 61 (2005) 2505.
[8] M. Grazul, E. Budzisz, Coord. Chem. Rev. 253 (2009) 2588.
[9] B.S. Creaven, M. Devereux, D. Karcz, A. Kellett, M. McCann, A. Noble, M. Walsh,
J. Inorg. Biochem. 103 (2009) 1196.
[10] A. Kulkarni, S.A. Patil, P.S. Badami, Eur. J. Med. Chem. 44 (2009) 2904.
[11] M.P. Sathisha, U.N. Shetti, V.K. Revankar, K.S.R. Pai, Eur. J. Med. Chem. 43 (2008)
2338.
[12] K.B. Gudasi, M.S. Patil, R.S. Vadavi, Eur. J. Med. Chem. 43 (2008) 2436.
[13] G.B. Bagihalli, P.G. Avaji, S.A. Patil, P.S. Badami, Eur. J. Med. Chem. 43 (2008)
2639.
[14] B.S. Creaven, D.A. Egan, D. Karcz, K. Kavanagh, M. McCann, M. Mahon, A. Noble,
B. Thati, M. Walsh, J. Inorg. Biochem. 101 (2007) 1108.
 B.S. Creaven et al. / Spectrochimica Acta Part A 84 (2011) 275–285 285

[15] B.S. Creaven, D.A. Egan, K. Kavanagh, M. McCann, M. Mahon, A. Noble, B. Thati, [26] P.J. Hay, J. Chem. Phys. 66 (1977) 4377.
M. Walsh, Polyhedron 24 (2005) 949. [27] K. Raghavachari, G.W. Trucks, J. Chem. Phys. 91 (1989) 1062.
[16] B.S. Creaven, D.A. Egan, K. Kavanagh, M. McCann, A. Noble, B. Thati, M. Walsh, [28] C. Lee, W. Yang, R.G. Parr, Phys. Rev. B37 (1988) 785.
Inorg. Chim. Acta 359 (2006) 3976. [29] A.D. Becke, J. Chem. Phys. 98 (1993) 5648.
[17] B. Thati, A. Noble, R. Rowan, B.S. Creaven, M. Walsh, M. McCann, D. Egan, K. [30] Available from http://www.chemcraftprog.com.
Kavanagh, Toxicol. in Vitro 21 (2007) 801. [31] R. Bauernschmitt, R. Ahlrichs, Chem. Phys. Lett. 256 (1996) 454.
[18] B. Thati, A. Noble, B.S. Creaven, M. Walsh, M. McCann, K. Kavanagh, M. Devereux, [32] A. Dreuw, J.L. Weisman, M. Head-Gordon, J. Chem. Phys. 119 (2003) 2943.
D.A. Egan, Cancer Lett. 250 (2007) 128. [33] C. Daniel, Coord. Chem. Rev. 238 (2003) 143.
[19] B. Thati, A. Noble, B.S. Creaven, M. Walsh, M. McCann, K. Kavanagh, M. Devereux, [34] V. Barone, F. Fabrizi de Biani, E. Ruiz, B. Sieklucka, J. Am. Chem. Soc. 123 (2001)
D.A. Egan, Cancer Lett. 248 (2007) 321. 10742.
[20] I. Georgieva, N. Trendafilova, B.S. Creaven, M. Walsh, A. Noble, M. McCann, [35] J.P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 78 (1997) 1396.
Chem. Phys. 365 (2009) 69. [36] J. Preat, D. Jacquemin, V. Wathelet, J.-M. André, E.A. Perète, J. Phys. Chem. A 110
[21] Tz. Mihaylov, N. Trendafilova, I. Kostova, I. Georgieva, G. Bauer, Chem. Phys. (2006) 8144.
327 (2006) 209. [37] W. Zhao, Y. Ding, Q. Xia, J. Comp. Chem. 32 (3) (2011) 545.
[22] I. Georgieva, N. Trendafilova, W. Kiefer, V.K. Rastogi, I. Kostova, Vib. Spectrosc. [38] M.T. Cancès, B. Mennucci, J. Tomasi, J. Chem. Phys. 107 (1997) 3032.
44 (2007) 78. [39] M. Cossi, V. Barone, B. Mennucci, J. Tomasi, Chem. Phys. Lett. 286 (1998) 253.
[23] A. Karaliota, O. Kretsi, C. Tzougraki, J. Inorg. Biochem. 84 (2001) 33. [40] B. Mennucci, J. Tomasi, J. Chem. Phys. 106 (1997) 5151.
[24] M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A. Robb, J.R. Cheeseman, [41] F.A. Cotton, G. Wilkinson, C.A. Murillo, M. Bochmann, Advanced Inorganic
G. Scalmani, V. Barone, B. Mennucci, G.A. Petersson, H. Nakatsuji, M. Caricato, Chemistry, 6th ed., John Wiley & Sons, New York, 1999.
X. Li, H.P. Hratchian, A.F. Izmaylov, J. Bloino, G. Zheng, J.L. Sonnenberg, M. Hada, [42] U. Muller, Inorganic Structural Chemistry, Wiley, Chichester, 1995.
M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, [43] B. Żurowska, A. Brzuszkiewicz, Polyhedron 27 (2008) 1623.
O. Kitao, H. Nakai, T. Vreven Jr., J.A. Montgomery, J.E. Peralta, F. Ogliaro, M. [44] K. Deka, N. Barooah, R.J. Sarma, J.B. Baruah, J. Mol. Struct. (Theochem) 827 (2007)
Bearpark, J.J. Heyd, E. Brothers, K.N. Kudin, V.N. Staroverov, R. Kobayashi, J. Nor- 44.
mand, K. Raghavachari, A. Rendell, J.C. Burant, S.S. Iyengar, J. Tomasi, M. Cossi, N. [45] P. Kopel, Z. Trávnícek, J. Marek, J. Mrozinski, Polyhedron 23 (2004) 1573.
Rega, N.J. Millam, M. Klene, J.E. Knox, J.B. Cross, V. Bakken, C. Adamo, J. Jaramillo, [46] M.A. Kiskin, I.G. Fomina, G.G. Aleksandrov, A.A. Sidorov, V.M. Novotortsev, Y.V.
R. Gomperts, E.E. Stratmann, O. Yazyev, A.J. Austin, R. Cammi, C. Pomelli, J.W. Rakitin, Z.V. Dobrokhotova, V.N. Ikorskii, Y.G. Shvedenkov, I.L. Eremenko, I.I.
Ochterski, R.L. Martin, K. Morokuma, V.K. Zakrzewski, G.A. Voth, P. Salvador, Moiseev, Inorg. Chem. Commun. 8 (2005) 89.
J.J. Dannenberg, S. Dapprich, A.D. Daniels, Ö.J. Farkas, B. Foresman, J.V. Ortiz, J. [47] I. Georgieva, N. Trendafilova, A. Aquino, H. Lischka, J. Phys. Chem. A 109 (2005)
Cioslowski, D.J. Fox, Gaussian 09, Revision A.1, Gaussian, Inc., Wallingford, CT, 11860.
2009. [48] K. Nakamoto, Infrared and Raman Spectra of Inorganic and Coordination Com-
[25] A.J.H. Watchers, J. Chem. Phys. 52 (1970) 1033. pounds, Part III, Wiley, New York, 1978.